Server - Few enhancements on tasks

Hello,

Would it be possible to enhance the server with the following functionalities for tasks management?

Add a button to rerun all failed tasks (or, even better, a checkbox allowing users to select specific tasks to rerun, with "Select All/None" options).

Enable minimize/maximize groups for better organization.

Create task dependencies to define execution order and relationships.

The first feature would be particularly helpful in scenarios of general failures (e.g., server downtime, emails blocked by Exchange). The second would improve page readability, and the third—most critical for us—would provide a clear overview of the extraction chain's process flow, including what happens and when. Currently, this is managed at the project level, but you need to delve into each project to see the details of what is loaded and when.

Best regards,
Chris

1 Like

Hello,

In order to manage failures, generally we set our tasks as cyclic and we have a "front" layer used by each projet to control if the task has already done what is necessary during the day. In that case we just exit.

About the logs, you can use the status action in each project to output whatever information you want. It will be available in the journal log (although truncated, I must admit it's hard to accept) but you can also output to a log file. Remember that one task can run another using the Easymorph server command action.

This is already on the middle-term roadmap. Next year, we will introduce pipelines - a new feature that allows building data pipelines from tasks. It will do exactly that.

The pipelines will only be available in the Enterprise edition.

2 Likes

Hi RJO,

Thank you for your message. Indeed, I’m aware of the available workarounds, and we already use log files to monitor our server. For instance, we have a project that runs every minute to log the available RAM, and another that calculates task durations. These metrics are then visualized in a Power BI report using Gantt charts and line charts to identify overlapping schedules and potential memory risks.

That said, in the event of a general failure, it could be very useful to have a button that relaunches all projects currently in error.

I also find your idea of cyclic tasks excellent, and we will explore how to implement something similar in our existing projects.

Thanks again for your input.

Best regards,
Chris